Caregiver reports and a few studies of older immigrants with Alzheimer's dementia suggest a reversion to the first language in the course of the disease. This proposed 12 month study investigates this reversion to the first language and its affect on the retrieval of autobiographical memories among older, bilingual, Hispanic immigrants in the early stages of Alzheimer's dementia (AD). There are three hypotheses in the study. First, older immigrant Hispanic patients with minimal to mild AD will show a consistent reversion to the first language (Spanish). Second, memories recalled in Spanish will be more specific, detailed, and emotionally salient than in English. Third, facilitating specific retrievals in Spanish will serve to cue additional recollections whereas retrievals in English will not. Theoretically, the study will show that language-specific retrieval serves to assist memory search abilities damaged in early AD by increasing cuespecificity. Practically, the study advocates use of the first language where reminiscence therapy is used in treating older bilingual patients with Alzheimer's dementia. The study will be conducted at the Alzheimer's Disease Center at Northwestern University with older j bilingual Hispanic immigrants with AD (minimal & mild) and matched healthy controls. Reversion to the first language will be assessed using the Bilingual Aphasia Test and subtests of standard aphasia batteries. Autobiographical memory will be tested by cuing memories in one language and then cuing those same memories on a subsequent day in the other language, and then reversing the order of languages in another pair of sessions. The study will serve as a pilot project for an R0-1 application that will (1) identify the specific language abilities lost earlier in the second language and preserved in the first and (2) examine how particular language dimensions (phonological? morphological? syntactic? prosodic?) act to increase the specificity of the cue and optimize memory search.